Frazier is an assistant professor at the Crump College of Law at St. Thomas University and a Tarbell fellow.

The first — and really only — order of business for the government is to solve problems beyond the grasp of a single person or a small community. In exchange for that service, we the people surrender some of our income and liberty. This grand bargain breaks down when the government decides it’s got other things to do besides take care of everything from our sewage to our space debris.

The longer the government falls short of our expectations, the more likely the people will be to opt out of their own obligations, such as voting. This dangerous tit-for-tat is hard to reverse. A less effective government sparks a less dutiful public, which makes it harder for the government to perform, and so on.

That’s precisely why the first agenda item for every politician who wins in November ought to be basic government work. For Congress, it’s time to get serious about regulating artificial intelligence. A series of performative hearings in which the celebrities of AI, such as Sam Altman, made flashy appearances on Capitol Hill have yet to generate meaningful legislation. The AI Policy Roadmap recently released by Senate Majority Leader Chuck Schumer (D-N.Y.) and three other senators did little to assuage concerns that Congress is happy to let states lead on AI governance.

A handful of small bills might address specific AI issues, but — absent passage of comprehensive legislation — the public is right to question if lawmakers are taking their part of the deal seriously.

For the president, the relative tranquility of the present is the proper time to prepare for future, inevitable emergencies. The federal government’s response to Covid-19 was far from perfect. From the initial confusion over who should wear a mask and when, to prolonged and troubling efforts to quash debate over the proper response, there is tremendous room for improvement.

The nature of our interconnected and global world means that another crisis will unfortunately emerge sooner than later. Imagine the good that would come about from the president overseeing massive emergency drills in preparation of threats like the deployment of a bioweapon. Picture how much smoother the Covie response would have been if the American people had preexisting stores of emergency supplies. These efforts are flashy but they are what’s required by an effective government.

For the courts, the integration of AI into our legal system should not be left to chance. State and federal courts alike have clogged dockets and insufficient support for individuals who cannot afford a lawyer but also cannot afford to lose whatever claim brought them before the court. AI will not and cannot entirely solve those issues, but the technology should be used to the greatest extent possible to help those who too commonly find that our legal system fails to provide the justice it promises.

None of these initiatives are partisan. And all of them go to the purpose of the government: solving problems. There’s a long list of similar, basic projects that would go a long way toward making the government more effective and efficient. The longer this grows, the greater the odds of (even more) discontent and distrust. That’s why candidates across the spectrum should focus less on short-term, shiny projects and much more on what will make American lives easier both now and for decades to come.

The latter kind of project won’t get headlines on Fox or MSNBC, but it will keep the wheels of government turning, which is no small feat given decades of rust.